René Laqua

Prior Shape Level Set Segmentation on Multistep Generated Probability Maps of MR Datasets for Fully Automatic Kidney Parenchyma Volumetry

Fully automatic 3-D segmentation techniques for clinical applications or epidemiological studies have proven to be a very challenging task in the domain of medical image analysis. 3-D organ segmentation on magnetic resonance (MR) datasets requires a well-designed segmentation strategy due to imaging artifacts, partial volume effects, and similar tissue properties of adjacent tissues. We developed a 3-D segmentation framework for fully automatic kidney parenchyma volumetry that uses Bayesian concepts for probability map generation. The probability map quality is improved in a multistep refinement approach. An extended prior shape level set segmentation method is then applied on the refined probability maps. The segmentation quality is improved by incorporating an exterior cortex edge alignment technique using cortex probability maps. In contrast to previous approaches, we combine several relevant kidney parenchyma features in a sequence of segmentation techniques for successful parenchyma delineation on native MR datasets. Furthermore, the proposed method is able to recognize and exclude parenchymal cysts from the parenchymal volume. We analyzed four different quality measures showing better results for right parenchymal tissue than for left parenchymal tissue due to an incorporated liver part removal in the segmentation framework. The results show that the outer cortex edge alignment approach successfully improves the quality measures.

Normal Breast Parenchyma: Contrast Enhancement Kinetics at Dynamic MR Mammography—Influence of Anthropometric Measures and Menopausal Status

PURPOSE To study T1 baseline signal intensity (SI) and contrast material enhancement kinetics of normal breast parenchyma by using dynamic contrast-enhanced (DCE) magnetic resonance (MR) mammography and to determine the influence of anthropometric measures and menopausal status on the variability of these features. MATERIALS AND METHODS Institutional review board approval and written informed consent were obtained. Between June 2008 and September 2011, 345 women (age range, 26-81 years; mean age, 51.3 years ± 11.6 [standard deviation]) underwent DCE MR mammography, with T1-weighted three-dimensional MR images (repetition time msec/echo time msec, 8.86/4.51; flip angle, 25°) acquired with a 1.5-T whole-body MR unit before and 1, 2, 3, 4, and 5 minutes after a gadobutrol bolus injection of 0.1 mmol per kilogram of body weight. Regions of interest were traced manually, and T1 SI of parenchyma was recorded. The influence of different predictors of T1 baseline SI and contrast enhancement was studied by using random-effects models. RESULTS T1 baseline SI varied considerably between women, with a mean of 167.7 ± 49.2 (71.4-424.7 [range]) and 175.9 ± 48.9 (51.8-458.3) in the right and the left breast, respectively (P < .01). T1 baseline SI increased linearly with age (P < .0001) and body weight (P < .0001). After contrast material delivery, relative percentage of enhancement was 8.1%, 13.8%, 18.2%, 22.1%, and 24.6% at 1, 2, 3, 4, and 5 minutes, respectively, but varied considerably between women. Contrast enhancement was 9.3% in the lowest quintile and 47.4% in the highest. Contrast enhancement increased with body weight (P < .01) but decreased in postmenopausal women (P < .01). Women with higher baseline T1 SI tended to have a higher contrast enhancement slope. CONCLUSION Anthropometric measures and menopausal status contribute to a large variability in contrast enhancement of normal breast parenchyma. This might influence the interpretation of contrast enhancement kinetics of breast lesions and current strategies for determining contrast medium dose for breast MR imaging.

An efficient level set method for simultaneous intensity inhomogeneity correction and segmentation of MR images.

Intensity inhomogeneity (bias field) is a common artefact in magnetic resonance (MR) images, which hinders successful automatic segmentation. In this work, a novel algorithm for simultaneous segmentation and bias field correction is presented. The proposed energy functional allows for explicit regularization of the bias field term, making the model more flexible, which is crucial in presence of strong inhomogeneities. An efficient minimization procedure, attempting to find the global minimum, is applied to the energy functional. The algorithm is evaluated qualitatively and quantitatively using a synthetic example and real MR images of different organs. Comparisons with several state-of-the-art methods demonstrate the superior performance of the proposed technique. Desirable results are obtained even for images with strong and complicated inhomogeneity fields and sparse tissue structures.

A level set based framework for quantitative evaluation of breast tissue density from MRI data.

Breast density is a risk factor associated with the development of breast cancer. Usually, breast density is assessed on two dimensional (2D) mammograms using the American College of Radiology (ACR) classification. Magnetic resonance imaging (MRI) is a non-radiation based examination method, which offers a three dimensional (3D) alternative to classical 2D mammograms. We propose a new framework for automated breast density calculation on MRI data. Our framework consists of three steps. First, a recently developed method for simultaneous intensity inhomogeneity correction and breast tissue and parenchyma segmentation is applied. Second, the obtained breast component is extracted, and the breast-air and breast-body boundaries are refined. Finally, the fibroglandular/parenchymal tissue volume is extracted from the breast volume. The framework was tested on 37 randomly selected MR mammographies. All images were acquired on a 1.5T MR scanner using an axial, T1-weighted time-resolved angiography with stochastic trajectories sequence. The results were compared to manually obtained groundtruth. Dices Similarity Coefficient (DSC) as well as Bland-Altman plots were used as the main tools for evaluation of similarity between automatic and manual segmentations. The average Dices Similarity Coefficient values were and for breast and parenchymal volumes, respectively. Bland-Altman plots showed the mean bias () standard deviation equal for breast volumes and for parenchyma volumes. The automated framework produced sufficient results and has the potential to be applied for the analysis of breast volume and breast density of numerous data in clinical and research settings.

Fast Implementations of the Levelset Segmentation Method With Bias Field Correction in MR Images: Full Domain and Mask-Based Versions

Intensity inhomogeneity represents a significant challenge in image processing. Popular image segmentation algorithms produce inadequate results in images with intensity inhomogeneity. Existing correction methods are often computationally expensive. Therefore, efficient implementations for the bias field estimation and inhomogeneity correction are required. In this work, we propose an extended mask-based version of the levelset method, recently presented by Li et al. [1]. We develop efficient CUDA implementations for the original full domain and the extended mask-based versions. We compare the methods in terms of speed, efficiency, and performance. Magnetic resonance (MR) images are one of the main application in practice.

Fully Automated Renal Tissue Volumetry in MR Volume Data Using Prior-Shape-Based Segmentation in Subject-Specific Probability Maps

Organ segmentation in magnetic resonance (MR) volume data is of increasing interest in epidemiological studies and clinical practice. Especially in large-scale population-based studies, organ volumetry is highly relevant requiring exact organ segmentation. Since manual segmentation is time consuming and prone to reader variability, large-scale studies need automatic methods to perform organ segmentation. In this paper, we present an automated framework for renal tissue segmentation that computes renal parenchyma, cortex, and medulla volumetry in native MR volume data without any user interaction. We introduce a novel strategy of subject-specific probability map computation for renal tissue types, which takes inter- and intra-MR-intensity variability into account. Several kinds of tissue-related 2-D and 3-D prior-shape knowledge are incorporated in modularized framework parts to segment renal parenchyma in a final level set segmentation strategy. Subject-specific probabilities for medulla and cortex tissue are applied in a fuzzy clustering technique to delineate cortex and medulla tissue inside segmented parenchyma regions. The novel subject-specific computation approach provides clearly improved tissue probability map quality than existing methods. Comparing to existing methods, the framework provides improved results for parenchyma segmentation. Furthermore, cortex and medulla segmentation qualities are very promising but cannot be compared to existing methods since state-of-the art methods for automated cortex and medulla segmentation in native MR volume data are still missing.

A fast global variational bias field correction method for MR images

Magnetic resonance (MR) images are prone to inhomogeneity artefacts that hinder an efficient automatic segmentation. Existing correction methods are often dependent on initialization and computationally expensive. This paper proposes a novel variational approach for the simultaneous bias field correction and image segmentation together with its efficient implementation, which produces the global solution that does not depend on initializations. The method is compared against another recently proposed method in terms of speed, efficiency, and performance.

Pharynx Segmentation from MRI Data for Analysis of Sleep Related Disoders

In our project, soft tissue structures of a throat are examined via MRI and anatomic risk factors for sleep related disorders are studied. Segmentation of pharyngeal structures is the first step in three dimensional analysis of throat tissues. We present a pipeline for pharynx segmentation with semi-automatic initialization. The automatic part of the approach consists of three steps: smoothing, thresholding, and 2D and 3D connected component analysis. Whereas two first steps are rather common, the third step provides a set of general rules for extraction of the pharyngeal component. Our method is minimally interactive and requires less than one minute to extract the pharyngeal structures, including the operator interaction part. The approach is evaluated qualitatively using 6 data sets by measuring volume fractions and the Dices coefficient.

Lung Segmentation of MR Images: A Review

Magnetic resonance imaging (MRI) is a non-radiation based examination method, which gains an increasing popularity in research and clinical settings. Manual analysis of large data volumes is a very time-consuming and tedious process. Therefore, automatic analysis methods are required. This paper reviews different methods that have been recently proposed for automatic and semi-automatic lung segmentation from magnetic resonance imaging data. These techniques include thresholding, region growing, morphological operations, active contours, level sets, and neural networks. We also discuss the methodologies that have been utilized for performance and accuracy evaluation of each method.

Automatic Pharynx Segmentation from MRI Data for Analysis of Sleep Related Disorders

In our project, we analyse throat structures using magnetic resonance imaging (MRI) to associate anatomic risk factors with sleep related disorders. Pharynx segmentation is the first step in the three-dimensional analysis of throat tissues. We present a pipeline for automatic pharynx segmentation. The automatic part of the approach consists of three steps: smoothing, thresholding, 2D and 3D connected component analysis. Whereas two first steps are rather common, the third step provides a set of general rules for the automatic extraction of the pharyngeal component. Our method requires less than one minute to extract the pharyngeal structures. The approach is evaluated quantitatively on 30 data sets using region-based and edge-based measures.